AI engineered for precision, speed, and reliability.
Phase 1
Refine the query
In order to optimize the accuracy of an answer that an LLM generates, the inputs the LLM receives must be refined for comprehension. The clearer and more understandable the query, the better the output.
Specification
- 1.1Check for safety and relevance
- 1.2Optimize query comprehension
- 1.3Check for Workflows automation
- 1.4Check for Custom Answer
Phase 2
Retrieve relevant content
The Fin AI Engine™ retrieval process, powered by our proprietary custom fin-cx-retrieval model, searches across all available knowledge sources (help center articles, resolved conversations, structured snippets, integrated systems etc) and selects the most relevant information to be used to answer accurately.
Specification
- 2.1Understand intent behind the question
- 2.2Search across all available content sources
- 2.3Match on semantics, not just keywords
Phase 3
Rerank for precision
The Fin AI Engine™ reranking process, powered by our proprietary custom fin-cx-reranker model, takes the retrieved content and scores each piece for relevance, accuracy, and usefulness in context. It then selects the final piece(s) for the LLM to use.
Specification
- 3.1Score relevance with respect to query
- 3.2Evaluate context match and resolution fit
- 3.3Downrank outdated or low-confidence sources
- 3.4Output: final selected content for generation
Phase 4
Generate a response
Once a query has been checked and optimized, the next stage is to generate a response using the LLM. For this task, the Fin AI Engine™ uses a bespoke and enhanced generative process, hand-engineered and tested at scale to generate answers with the highest chance of resolving customer questions. At this point, any custom Guidance provided by the user to control the generated answer is also applied. This is the final step of our Retrieval Augmented Generation system (RAG for short).
Specification
- 4.1Integrate and augment
- 4.2Interpret and apply Guidance
- 4.3Generate response
Phase 5
Validate accuracy
In the final step of the process, the Fin AI Engine™ performs checks to understand whether the output from the LLM meets the necessary response accuracy and safety standards.
Specification
- 5.1Validate the response
- 5.2Respond to customer
always-on
Engine optimization
To calibrate and enhance engine performance, the Fin AI Engine™ has advanced integrated tools that help optimize answer generation, efficiency, precision, and coverage.
Specification
- 6.1Fin customization and control
- 6.2AI analytics and reporting
- 6.3AI recommendations
safeguarding
AI Trust & Security
Intercom has implemented state-of-the-art security measures to protect Fin against a wide range of LLM threats, including those identified by the OWASP LLM Top 10. By consistently testing a variety of high-end LLMs, and deploying rigorous internal controls, security protocols, and safeguards, Fin is able to achieve the highest level of security and reliability while avoiding potential limitations and threats.
Specification
- 7.1Fin AI Security
- 7.2Regional hosting
- 7.3Compliance: International Standards
- 7.4Third-party AI Providers
FAQs
FAQs
How does Fin's AI Engine work?▼
Fin’s AI Engine™ is engineered for precision, speed, and reliability—powering the highest-performing and safest AI Agent in customer service.
It starts by refining each query—the inputs received must be refined for comprehension. This ensures every customer message is optimized for meaning and context. The Fin AI Engine™ then retrieves relevant content with our proprietary fin-cx-retrieval model and scores it for relevance, accuracy, and usefulness, selecting the best information to use. Next, it generates a response through a process tested at scale to generate answers with the highest chance of resolving customer questions. Finally, it performs checks to ensure the output meets the necessary accuracy and safety standards.
Fin’s AI Engine™ continuously optimizes itself with integrated tools that improve efficiency, precision, and coverage.
What makes the Fin AI Engine the highest performing?▼
The Fin AI Engine™ is a patented architecture purpose-built for customer service. Every layer is optimized for accuracy, speed, safety and reliability, so Fin resolves more queries and delivers higher-quality experiences than any other AI agent.
Is the Fin AI Engine ready for enterprise scale?▼
Absolutely. Fin is proven across millions of customer conversations, delivering 99.97% uptime, enterprise-grade reliability, and compliance with SOC 2, ISO 42001, and HIPAA standards.
